Measuring Up: Assessment in Collections and Technical Services

4/4/2003
Hogan Center, College of the Holy Cross
Worcester, MA

Morning speaker: Sheila Intner, Professor Emerita, Graduate School of Library and Information Science, Simmons College:
Making Your "Collections Work" Work for You

Breakout sessions:
Peter Spitzform (University of Vermont): Offsite Storage: Planning the Move is the Key
Beth Mazin (Memorial Hall Library, Andover, Mass.): Weeding: I Know Should But...
Peggie Partello (Keene State College) and Glenda Lins (OCLC): A Case of Ample Statistics: Collection Analysis with ACAS
Marianne Burke (Dana Medical Library, University of Vermont): Using LibQual to Improve Customer Services

Lunchtime speaker: Ann Caldwell (Brown University):
OLAC: Online Audiovisual Catalogers

Afternoon speaker: Sally Grucan (Head Cataloging Librarian, Wesleyan University Library):
Cataloging Statistics with a Purpose

Sheila Intner, Professor Emerita, Graduate School of Library and Information Science, Simmons College
Making Your "Collections Work" Work for You

Dr. Sheila Intner's keynote address was devoted to the important topic of library collection evaluation. She approached the subject as a "debunker," describing ten persistent myths and countering them with the realities she has learned from her years of teaching, consultation, and experience as a practicing librarian. (NOTE: Dr. Intner will publish an expanded version of her remarks in a forthcoming issue of Library Collections, Acquisitions, & Technical Services.)

Myth #1: "No one needs to do inventories-no one has done them for years & library service hasn't halted!" Dr. Intner noted that shelf-reading has also been given up in many larger libraries, and accessions lists are frequently things of the past. Nevertheless, a collection evaluation must begin with a collection profile, for which an inventory is essential. Three corollaries accompany this point:
Corollary #1: "I 'know' what I have without an inventory." While an intuitive sense of one's collection can be a useful supplement, it is dangerous to completely trust it.
Corollary #2: "I 'know' what people use without data to tell me." While it is true that data can lie, this will not be true if the data is gathered properly and manipulated appropriately. Anecdotal evidence, by contrast, is often wrong.
Corollary #3: "I don't need statistics to tell me what is going on in my library!" On the contrary, our direct experience cannot provide us with the whole picture. Information which slips beyond our "peripheral vision" is important, even in small increments, to an understanding of the collection and how it is used.

Myth #2: "Strategic planning for collections is much more important than inventorying them." Strategic planning is indeed of great importance, and inventorying is an essential part of that activity. Conducting an inventory helps librarians gain an intimate, profound knowledge of the makeup and breakdown of a collection. Strategic planning is about controlling the process of change, but in order to do that you must know where your beginning point is.

Myth #3: "Every library is individual and you can't compare it to any other library." Each library does have its individuality. But are the differences between libraries so profound that they prevent comparisons? Comparisons do not have to be made between identical entities in order to be useful. There may indeed be a few libraries, such as the Library of Congress, which are actually incomparable to any others, but the example itself shows how rare the case is. (To illustrate this point, Dr. Intner asked all of the attendees who considered themselves blondes to stand up, to the amusement of the group. It was clear that, although everyone who stood was a blonde, there were many different hair colors in evidence.) Comparisons can profitably be made between the collections of peer libraries, in terms of variables such as relative strength and currency. Identification with libraries in a peer group can provide the motivation for change in a collection, but data (particularly inventory data) is still required to quantify the need for change.

Myth #4: "'Good' is too subjective to express in substantive terms." Unfortunately, subjective judgments can't all be equally right when they're in conflict. There must be a means of operationalizing concepts such as "good," particularly in the current environment, in which assessment is more important than ever. There is an even greater danger here, in that the opposition to evaluation can extend to performance and services as well as collections. The resistance to accountability in turn can alienate funding bodies. As Dr. Intner emphasized, even "motherhood and apple pie" can be evaluated, since not all mothers and apple pies are equally good!

Corollary #4: "My collection is good even though it has some problems." This may be true, but again it depends on how "good" is defined. Fortunately, this can in fact be operationalized. Dr. Intner provided a set of suggested, reasonable targets, which may be used to define a good collection. Some examples are:

• 75% of requested materials can be found on the day they are requested; at least 10% of the rest within 1 week; at least 10% of the remainder within 2 weeks (no more than 5% of requests are undeliverable within 2 weeks);
• 90% of all reference questions can be answered without going beyond local resources/access;
• 90% of the works identified as "best sellers" in major newspapers and television/radio talk shows are purchased, in sufficient numbers that patrons do not wait more than 4 months to fill a request;
• important websites in high interest areas and/or departmental subjects are easily accessed through the library's catalog or local website.

By contrast, a bad collection may exhibit characteristics such as:

• The average age of collections is older than 10 years in the sciences and technology, and older than 15 years in the social sciences;
• purchasing is linked to factors such as tradition, not relevance or expected use;
• illustrations and other visual images in books and other children's materials aren't in color.

Although there are no current national standards in this area, these points suggest the possibilities for objective evaluation of the quality of a collection.

Myth #5: "Bigger is better when it comes to collections." The psychological appeal of having lots of things is strong, and of course it is very difficult to actually do weeding in full view of the public during the day. But browsing through collections of little-used, shabby and dusty collections cause fatigue and discouragement for library patrons. In short, more of what patrons want is better, but more of what they don't want is not.

Myth #6: "My library can buy what our patrons need, and uses ILL only for odds and ends." Collection development arose as a distinct specialty largely because this is not the case. The gap between what almost all libraries actually need, and what they can afford to buy, has been growing for fifty years. Making the claim that we can buy all we need may also have the political consequence of become more vulnerable to budget cuts. Again, assessing what a library actually has, via an inventory, will help provide a clear picture of what that library in fact needs.

Myth #7: "I can't inventory digital materials because the library [select from the following]: a) doesn't own them; b) doesn't store them physically on the shelves; c) can't determine what's in the services we buy; d) can't tell what's being used; e) all of the above." Although difficult, it is possible to inventory digital materials. The distinction between things stored on the shelves, and things accessed only via computer is becoming invisible, to both patrons and funders. The concept of "holdings" in relation to paid-for (purchased or licensed) electronic resources has changed somewhat, but one still can provide a count of access to online services. It is understood that, at this time, vendors may still have difficulty providing this information; librarians should lobby their vendors to get the kinds of information they really need.

Myth #8: "Collecting lots of data is a useless, thankless task." A well-planned data collection process will include preliminary work to determine just what is needed, and for which purposes. This in itself will reduce the "thanklessness" of the process. Pertinent data is essential in order to justify claims about a library's operations, and to provide background for obtaining needed resources. Good data can enhance one's knowledge, vision, credibility and status within an institution. To put it another way, while knowing "the numbers" doesn't guarantee that you'll win your battles, not knowing the numbers guarantees that you'll lose.

Myth #9: "No one wants to see all that evaluation data." It is true that nobody will want to see your worksheets! But evaluation data can be a powerful tool if presented properly. A good breakdown will follow this format: an executive summary, followed by an introduction with background information; the collection profile (a/k/a inventory data at summary level) broken down by subject or department and also by media; comparisons of selected features with peer libraries when available; reports of quality measures by department; conclusions about collection performance, and recommendations. Appendices may include a bibliography of sources consulted, and potentially the raw data itself.

Myth #10: "I haven't got the time for statistics and all that mathematical stuff." There are several means available to maximize the return on your effort, and help you budget your time: 1) hire people to help you through the process of digesting data (not to do it for you); 2) delegate responsibility to someone who is interested in the work, and work closely with them; 3) form an evaluation team to share the work; 4) seek input on problems from specialists outside the library but within the parent institution. In addition, it's important to remember that basic office software has developed to the point of being of great assistance with these tasks.

In conclusion, Dr. Intner provided "Ten Commandments" as a complement to her ten myths:

1. Plan everything;
2. Train everybody;
3. Economize;
4. Count everything;
5. Use technology;
6. Organize data by subject;
7. Report understandably;
8. Use both narrative prose and visuals;
9. Summarize;
10. Let your "collections work" work for you!

Reported by David Miller
Curry College
NETSL Writer/Editor

Offsite Storage: Planning the Move is the Key
Peter Spitzform (University of Vermont)

Peter Spitzform, Collection Development Librarian at the Bailey/Howe Library of the University of Vermont, described his experience with planning the move of about 130,000 monographic volumes to a storage facility, called the Library Research Annex (LRA). Spitzform's presentation was particularly welcome at a time when many libraries are facing space problems and the topic of offsite depositories is not well covered in the professional literature.

Mr. Spitzform recommended to begin planning for an offsite storage before a library building reaches the functional capacity of 85%. In this presentation he identified four major planning elements: 1) Assessing space needs; 2) Selecting titles to be moved; 3) Organizing the move; and 4) Marketing the new facility.

1) Assessing space needs
Key to planning is an assessment of the collections' room for growth. In that phase libraries need to determine some goals: how much shelf-space needs to be freed-up to make the collections and the space manageable, and how many years of growth the Library anticipates gaining.

At UVM, Mr. Spitzform was able to answer several of these questions by using reports on collections usage and patterns of acquisition growth by Library of Congress class number. These were organized by Library of Congress classification and were used to project the rate of growth of the collection. Space calculations were based on an average of 10.8 monographs and roughly 8 volumes of bound journals per foot. These will vary from library to library.

2) Selecting titles to be moved
The selection criteria for the identification of these 130,000 titles were two: publication date prior to 1986, and no known circulation or browse. Lists of titles were created using these criteria. The next step was to make sure that the physical volumes were indeed available on the shelf and to mark those that could be moved. The lists were printed. In a first pass student assistants marked single copies; in a second pass library staff made decisions on multiple copies and multi-volume sets; for the later the decision was based on a 20% circulation rule. As a precautionary measure a third pass involved some checking from subject specialists to ensure no title that may be fundamental to a field of teaching or research was being overlooked.

3) Organizing the move
Based on this experience at UVM, Spitzform made a clear distinction between a library which is planning its first move to an offsite facility or moving a separate distinct collection, and one which is moving an additional collection to an existing storage facility as it was the case at UVM. UVM staff was facing an even more difficult challenge because it was adding to an existing universe of materials which was already housed and shelved by call number at the offsite storage facility. Other important points stressed by Spitzform included the availability of adequate staffing to facilitate the actual move and the offer of some form of incentive for the staff. At UVM staff who participated in the move received extra time off.

The process of selecting the mover began with the preparation of RFPs. After receiving various RFPs, a professional library mover was selected. An important factor in this decision was the experience and the special skills of the company. The UVM project did not involve a simple physical move of library materials; it involved inter-shelving materials in call number order.

The change of location code in the online catalog was scheduled after the titles had been inter-shelved and corresponding barcodes scanned with a hand-held barcode scanner. The data was downloaded and records were updated in batch mode to the new offsite storage location.

4) Marketing the new facility
Ensuring that users understand the purpose and function of the new facility is an important aspect of marketing the offsite storage depository. Only UVM staff has access to the storage area itself. The physical layout (narrow aisles and high shelves) precludes from having users browse through the shelving area. The offsite storage facility is currently open every day 1-5 pm. The staff is able to keep up with the demand for materials to circulate. The turn-around time to retrieve titles is one day, which is acceptable for most users. There are plans to add reference service to the facility.

As a result of the move of 130,000 monographs, UVM gained space for as many as 8 years of additional materials. The success of this project was partly due to the fact that it was based on real data and that this data was derived from collection usage, i.e. based on titles that had not circulated.

Reported by Dominique Coulombe
Brown University

Weeding: I Know Should But...
Beth Mazin (Memorial Hall Library, Andover, Mass.)

Valuable tips and information about weeding activities were exchanged in this spirited, often lighthearted, discussion among more than thirty participants, facilitated by Mazin, who, as Assistant Director at the Memorial Hall Library in Andover, is also responsible for collection development. Mazin began the session by distributing a thick handout to the group, which described the weeding plan developed and followed by staff at Memorial Hall Library, as well as the advantages and disadvantages of various weeding strategies. She quickly outlined her library's plan, explaining that several years ago the impetus for weeding had been to create space for new acquisitions and to reconfigure space for new formats of material. In addition, at that time, good usage statistics were becoming available through her consortium's online system, which bolstered a large-scale weeding effort with minimal staff involvement.

Then, having noted that most of the attendees at the conference were from academic libraries, Mazin asked if weeding were done differently in those libraries. Responders spoke about the involvement of faculty members in weeding decisions, strong reliance on circulation statistics, the need to determine in-house use, especially of non-circulating materials or of periodicals, the necessity of being aware of "core" materials and classics, the retention of "the last copy" within a library system, weeding more selectively in certain subject areas than in others, and conducting inventories prior to weeding. Questions arose about weeding gift materials and long-playing records. Of particular interest were disposal mechanisms for unwanted materials. These ranged from offering discards to Friends groups and other organizations, including prison libraries, to placing them in book sales, selling them to dealers, or relegating them to storage areas, to, finally, tossing them into dumpsters. As the session was drawing to a close, Mazin asked if anyone had experienced any controversial problems in weeding. None present had, and she drew one final roar of laughter and applause by saying, "If anybody goes to jail, I'll testify on your behalf!"

Memorial Hall Library's collection development policies, including its weeding guidelines, can be found at http://www.mhl.org/policies/cd.

Reported by Carol A. Hryciw-Wing
Rhode Island College

A Case of Ample Statistics: Collection Analysis with ACAS
Peggie Partello (Keene State College) and Glenda Lins (OCLC)

OCLC's ACAS (Automated Collection and Analysis Services) provides librarians with collection analysis, title overlap analysis, library collection comparisons (gap analyses) and list comparison reports. It also supports the WLN Conspectus and RLG Conspectus software. Information may be found at the demo site: http://acas.oclc.org. The guest user login provides a demo for NLM, Dewey and LC Classification systems.

OCLC can compare libraries to other libraries which use OCLC for their cataloging. ACAS will use two or more libraries that are selected for comparison reasons. Libraries may include as many libraries as they would like (and can afford) for this comparison but there has to be a minimum of two. The report does not include the names of the libraries that actually hold the item, but simply indicates that the title is available in one of the comparison libraries. One-on-one comparison cannot be done without permission. The report is broken down into several types. The match report includes titles that are in the requesting library collection and at least one of the comparison libraries. A close match report will include titles that may have a different author, date, edition, etc. The miss report includes titles that are in a comparison library but not in the requesting library. The statistical summaries will provide the number and percentage of matches, close matches and misses for each WLN Conspectus line. The WLN Conspectus has 24 subject divisions, 500 subject categories and 4000 descriptors. Reports may be limited by division, category, publication date, language, format, and audience.

The List Comparison Service compares a requesting library's collection with Books for College Libraries (3^rd ed), Choice's Outstanding Academic Titles and titles reviewed in Booklist. These reports may be used for collection development.

Libraries may use their own system data-LC or NLM call numbers. Records used for comparison will have to have call numbers, since the subject comparisons are based on the call number field (not on subject heading schemes). The library determines the date ranges used in the reports. In reports listing date ranges, information is included on records that do not have a date in the fixed field, so these titles may not be assigned a date. It is possible to compare different subject areas to different libraries. The requesting library does not have to compare the whole collection to another library. Public libraries may request reports that compare their branches. Consortia may be interested in overlap and uniqueness information. SuDocs numbers may be mapped to LC Classification. It is recommended that periodicals be excluded from the analysis due to the use of the first date, which will skew the age of the collection. Periodicals may be included in a secondary analysis. OCLC does use the libraries' archive records for analysis.

Currently the data set is a one-time analysis since the comparisons cannot be done online yet. The data is available for one year from the time of the report. It is possible to get the information on CD. This software works with Internet Explorer but not with Netscape. The cost is determined by the size of the collection and the level chosen (division vs. categories) and the number of libraries chosen for comparison.

Keene State College chose to use ACAS to evaluate their collections. They wanted to find out the strengths and weaknesses and the age of their collections. They chose two libraries for comparison, mainly due to cost. These two libraries were chosen from their peer institutions and from the public liberal arts colleges (COPLAC) listing.

The two major reports they received were the missing titles list and the age and content list. The missing title list includes the raw data-line number, title, author, edition, date, and publisher-for each title. This allowed them to compare their collection with two colleges that have similar programs.

The age and content list was done at the subject/titles level using the LC Conspectus. The dates were broken down by decades in the twentieth century until 1995After this date the information was done by individual year. The information provided does not tell when collections were weeded or when there were grants for collection purchases. They chose not to include their periodicals.

Keene State designed several reports from their data. They did a breakdown of their collection by audience-adult vs. juvenile. Other reports that provided valuable information were average number of items acquired by year, items per faculty, items for the last decade (by year), items per faculty in a discipline, and items per major. In education Keene State asked for a breakdown by categories within education. All this information showed them what collections needed to be updated, what collections needed more materials (education is their largest program but it does not have the largest collection), and what areas needed to be weeded (basically all areas).

There were some questions about where to include some aspects of the information. The subjects used were related to campus departments. Where do you include minors? Where do you include some departments that overlap subject areas (politics)? The number of registered students was not included in the evaluated data, just the majors. Keene State also recommends that the requesting library has the support to make decisions based on the results of the collection evaluation. Since these reports are expensive, it is recommended not to spend the money if the results cannot be used to make selection changes.

Reported by Karen Matthews
University of Vermont

Using LibQual to Improve Customer Services
Marianne Burke (Dana Medical Library, University of Vermont)

Marianne Burke, Director of the Dana Medical Library at the University of Vermont, described how that library implemented the ARL-supported LibQual+ survey and how they are using the results to improve users' satisfaction with library public and technical services. It soon became apparent that many people in the audience were either in the throes of a LibQual+ survey or were planning to embark on one in the near future. There were many questions about promoting the survey, the impact of the survey on staff, and the usefulness of the final report.

LibQual+ is a survey instrument and analysis service that is supported by the Association of Research Libraries. It has been adapted for library use from a business total market survey that not only assesses the level of satisfaction of users of a particular library, but also compares the level of user satisfaction of individual providers with others in the same cohort. A LibQual+ cohort is comprised of a group of similar libraries that all participate in the survey and analysis at the same time. The Dana Medical Library was in a cohort of other medical libraries, including some of the largest.

The survey instrument consists of 25 questions that have been gleaned over several trials. The questions cover access to information, effectiveness of service, how users feel about the library as a place, and patrons' feeling of personal control over their research activities. LibQual+ respondents are asked to assign a numerical value to their minimum service requirements, their desired service levels, and their perceptions of where on that scale their libraries' collections and services are located. The gap between the minimum service requirements and the user perception of service tells the libraries where they are succeeding, and where they need to improve. The desired service level is a goal that a library can strive to achieve. Even though getting the survey to every member of the library's constituency proved to be complicated, they had a response rate of 30%, which is excellent for a mailed survey. Ms. Burke mentioned that they might have received a better response if they had promoted the survey more heavily.

Dana Medical Library is using the detailed analysis of the survey that LibQual supplied to change systems in the library to improve customer service. For example, they improved their Web page to give patrons more guidance at that point and made improvements in the physical space, which ranked very low in the surveys. One change that had a very large impact was to deregulate food in the library, and they even offer snacks in the library during high-stress times. On a larger scale, they have improved their communication with the Medical School IT department in order to make them more responsive to user needs. They have also launched a series of discussions with the library constituents to address those problems that are not easily solved, such as perceived gaps in the collections. Through these discussions they hope to prioritize the users' needs and respond to the most pressing ones. They are also looking to other libraries to find solutions.

Ms Burke feels that the LibQual+ experience has been very positive for the Dana Medical Library, because they now have far more information about what they constituents want and expect, and how well they are living up to those expectations. As Ms. Burke wrote in her introduction to the session, LibQual+ has helped them to "get a handle on the only opinion that matters: that of the library users". They know what the users want; now they can work on providing it.

Reported by Christina Bellinger
University of New Hampshire

Ann Caldwell (Brown University)
OLAC: Online Audiovisual Catalogers

Since 2001, NETSL has provided a highlight of the work of a related library organization, as part of the Spring Meeting. Ann Caldwell was invited to describe the work of the Online Audivisual Catalogers (OLAC). Ms. Caldwell briefly reviewed the history of OLAC, including the farsighted librarians who served as its founders, the establishment of the OLAC Newsletter and discussion list, and the awards which the organization grants. OLAC sponsors a conference every other year: the next will be held from September 29 to October 3, 2004, in Montréal. The OLAC conference, which has a hands-on emphasis, features numerous workshops, larger group sessions with noted speakers, and group tours to AV resource-related sites. The conference features popular "Q&A" sessions on difficult and/or cutting-edge AV cataloging questions, which are reported in the Newsletter. OLAC also has regular meetings at ALA conferences. The Cataloging Policy Committee (CAPC), probably the most active group in OLAC, meets on Friday evenings at ALA. OLAC has prepared many valuable training materials, helped to develop Library of Congress Rule Interpretations and AACR2 rule revisions, and sponsored workshops. Ms. Caldwell also described OLAC's NACO AV Funnel Project, in which cooperating libraries contribute authority records to the national Name Authority File. Information about all of these activities, services and publications can be found at OLAC's web site.

Sally Grucan (Head Cataloging Librarian, Wesleyan University Library):
Cataloging Statistics with a Purpose

Sally Grucan, Head Cataloging Librarian at Wesleyan University, provided the NETSL attendees with a picture of what detailed yet informative cataloging statistics can be, using her institution's experience as the model. Her talk covered reasons for keeping statistics (good and bad), types and sources of statistical information, the uses to which statistics can be put, and sample presentations.

As background, as of Spring 2003 the Wesleyan University Library employed 17 librarians and 28 staff members. Twelve people work in technical services, and of these, four work in cataloging (three full-time and one part-time), which includes physical processing. There has been a loss of cataloging staff at Wesleyan, as is true for many other institutions: earlier, there were eight cataloging staff, including three professional catalogers. As partial compensation and to avoid backlogs, some work that was done in cataloging was moved into other departments, such as acquisitions, serials, scores and recordings, and special collections. Still, the workload for remaining cataloging staff is great. Since the phenomenon of more work handled by a smaller staff is hardly unique, many institutions could make use of detailed statistics for documenting staff workload, demonstrating the efficiency of operations, and building the case for potential additional staff.

Statistics can be useful tools for many purposes beyond documentation of workload. Other good reasons for keeping statistics include tracking cataloging costs, tracking vendor performance, answering specific questions, planning projects, supplying numbers needed by others in the organization, and tracking individual and group performance. (Poor reasons for keeping statistics include "just in case", and "because we've always done it.")

Wesleyan makes extensive use of outsourcing in its cataloging operations. This addresses two problems: the loss of staff time and expertise, and the need to move items through technical services quickly. The Library contracts with TechPro for difficult original cataloging and to replace the loss of foreign language expertise. Shelf-ready items are received through PromptCat (with YBP and Eastern as vendors), and LTI is used for authority processing.

It's not always, clear, however, that a department can outsource without significantly lowering cataloging quality or actually raising costs. Assembling the pertinent statistics and presenting them comprehensibly is essential in order to assess the success of an outsourcing operation. Ms. Grucan presented, via an extensive handout and slides, a suite of spreadsheets which bring together the pertinent information. The first sheet provides a summary of non-staff/staff/student cataloging costs for a given fiscal year (2002 in this instance). This is supported by detailed supplementary spreadsheets: non-staff cataloging costs (including PromptCat, TechPro, and physical processing by vendors); monthly OCLC costs by product code; PromptCat statistics showing acceptability of records received; staff cataloging costs by individual staff member and activity; and student worker costs. A final sheet projects OCLC costs into the next fiscal year.

This is a complex effort, but it results in the ability to track and compare costs at a fine level of granularity. In addition, it is very difficult to find useful or comparable statistics in the professional literature, as the variables examined in cataloging cost studies are too disparate. Since Wesleyan had already been keeping and using some types of statistics, they were able to gradually add to what they had.

The data involved comes from several sources, including OCLC system usage statistics, NELINET invoices, cumulative statistics from the Promptcat web site, and data from Wesleyan's library system. Some information is also compiled manually, including quality control statistics related to PromptCat and personal statistics kept by staff. Some statistics related to OCLC transactions are also kept manually, as a comparison with the data received from OCLC. In time, this redundant practice will be phased out if the two sources are consistently found to provide the same results.

Ms. Grucan continued her talk with a discussion of factors which have an impact on the balance between cataloging efficiency and quality, particularly in environments driven by shrinking budgets and staff. The great challenge is to define what "quality" denotes and to provide it at an acceptable level. This may require making a distinction between a more flexible acceptance of existing copy and continuing to hold original cataloging to a higher standard.

At Wesleyan, cataloging is provided at several levels, as deemed appropriate. Original cataloging uploaded to OCLC is performed at full level. The minimum for in-house copy cataloging is Core level or an approximation. PromptCat copy cataloging, which is intended to result in shelf-ready items, is treated as fast-track. Minimal level cataloging (MLC) may stand in temporarily until more extensive work can be done. Cataloger-designed templates assist cataloging, notably in special collections. Another factor with an impact on both efficiency and quality is a trend toward postcataloging assessment and cleanup, with some quality control operations performed after the fact. As mentioned, LTI provides authority control correction and updating. OCLC's Bibliographic Notification service provides updates to records used by Wesleyan, including added contents notes. Library system-generated reports are generated to signal conditions such as CIP records not upgraded, records without subject headings, and use of inappropriate item types. Student workers are trained for tasks such as correcting misspellings and fixing incorrect initial articles.

Ms. Grucan also touched on the effects of electronic resources and technology on the catalog, the cataloging process, and, inevitably, a share of the library budget. Among these are user appreciation of features such as single-line Google-type searching, tables of contents, jacket covers, and reviews. Developments such as XML as an encoding standard, and FRBR (IFLA's Functional Requirements for Bibliographic Records) as a data model, are having an impact on the balance between emphasis on detailed description of individual resources, and recording of data which helps clarify the relationships between and among resources. The shifting patterns of usage of electronic- vs. paper-based resources have implications not only for collection development but for what is cataloged and how. Products which aim to unify access to the traditional library OPAC and distributed electronic resources (such as Ex Libris' SFX product in Wesleyan's case) are worth serious consideration.

At the close of her talk Sally Grucan offered several "final thoughts:"
1) keep cost-related statistics;
2) reconsider copy cataloging;
3) don't shortchange unique materials;
4) be proactive (you'll look good);
5) consult with public services; and
6) outside influences and limited funds will change the catalog and cataloging in fundamental ways, "so go out and buy a good bottle of wine!"

Sally Grucan may be reached by email at sgrucan@wesleyan.edu for further information on the contents of her talk, including examples of the statistics discussed.

Reported by David Miller
Curry College
NETSL Writer/Editor